International Journal of Technology Assessment in Health Care

General Essays

Constructive Technology Assessment (CTA) as a tool in Coverage with Evidence Development: The case of the 70-gene prognosis signature for breast cancer diagnostics

Valesca P. Retèla1, Jolien M. Bueno-de-Mesquitaa1, Marjan J. M. Hummela2, Marc J. van de Vijvera3, Kirsten F. L. Doumaa4, Kim Karsenberga5, Frits S. A. M. van Dama6, Cees van Krimpena7, Frank E. Bellota8, Rudi M. H. Roumena9, Sabine C. Linna10 and Wim H. van Hartena10

a1 Netherlands Cancer Institute-Antoni van Leeuwenhoek Hospital

a2 University of Twente

a3 Academic Medical Center

a4 Netherlands Cancer Institute–Antoni van Leeuwenhoek Hospital

a5 Stichting DES-centrum

a6 Netherlands Cancer Institute–Antoni van Leeuwenhoek Hospital

a7 Kennemer Gasthuis Haarlem

a8 Spaarne Hospital Hoofddorp

a9 Máxima Medical Centre

a10 Netherlands Cancer Institute–Antoni van Leeuwenhoek Hospital


Objectives: Constructive Technology Assessment (CTA) is a means to guide early implementation of new developments in society, and can be used as an evaluation tool for Coverage with Evidence Development (CED). We used CTA for the introduction of a new diagnostic test in the Netherlands, the 70-gene prognosis signature (MammaPrint®) for node-negative breast cancer patients.

Methods: Studied aspects were (organizational) efficiency, patient-centeredness and diffusion scenarios. Pre-post structured surveys were conducted in fifteen community hospitals concerning changes in logistics and teamwork as a consequence of the introduction of the 70-gene signature. Patient-centeredness was measured by questionnaires and interviews regarding knowledge and psychological impact of the test. Diffusion scenarios, which are commonly applied in industry to anticipate on future development and diffusion of their products, have been applied in this study.

Results: Median implementation-time of the 70-gene signature was 1.2 months. Most changes were seen in pathology processes and adjuvant treatment decisions. Physicians valued the addition of the 70-gene signature information as beneficial for patient management. Patient-centeredness (n = 77, response 78 percent): patients receiving a concordant high-risk and discordant clinical low/high risk-signature showed significantly more negative emotions with respect to receiving both test-results compared with concordant low-risk and discordant clinical high/low risk-signature patients. The first scenario was written in 2004 before the introduction of the 70-gene signature and identified hypothetical developments that could influence diffusion; especially the “what-if” deviation describing a discussion on validity among physicians proved to be realistic.

Conclusions: Differences in speed of implementation and influenced treatment decisions were seen. Impact on patients seems especially related to discordance and its successive communication. In the future, scenario drafting will lead to input for model-based cost-effectiveness analysis. Finally, CTA can be useful as a tool to guide CED by adding monitoring and anticipation on possible developments during early implementation, to the assessment of promising new technologies.


The present CTA study and the RASTER study were financed by the Dutch Health Care Insurance Board. We are indebted to the women who participated in this study and to all co-workers in the 16 hospitals. Special acknowledgements to Laura van't Veer, Chad Gundy, Guus Hart, and Fiona Stewart for their input. W.H. van Harten is a non-remunerated member of the supervisory board of Agendia BV. M.J. van de Vijver is named inventor on a patent application for the 70-gene prognosis signature used in this study. All other authors declared no conflicts of interest.